My opinion is that people who only implement based on examples, without reading the actual protocol specification, have themselves to blame - and they are going to run into problems sooner or later, no matter how correct the exmaples are.
If people are too lazy to read 3261 I don't think there is anything IETF can do - or even should spend time on doing. Regards, Christer > -----Original Message----- > From: Thomas Froment [mailto:[EMAIL PROTECTED] > Sent: 13. joulukuuta 2007 10:43 > To: Jonathan Rosenberg > Cc: Christer Holmberg; [email protected] > Subject: Re: VS: [Sip] SIPit21: BNF future-proofing problem? > > Jonathan Rosenberg wrote: > > Correction, yes. Essential? Hardly. > > > > This is another question around essential corrections. I think they > > not only need to clearly represent a bug or a fix, but need > to be one > > that is important enough to merit documenting. Just like software, > > sometimes you don't ever bother fixing those sev 5s. Ever. > Just would like to raise a small warning with the "essential" > vs "non-essential" debate... Sometimes the most "costly" bugs > (in terms of interoperability problems) are the smallest stupid nits. > Just one example, the FIRST proxy example of the RFC 3665 is > broken... > Contact of subsequent request is not correctly pushed in the > Request URI with transport parameter, so, at SIPIT, I already > have seen many UAs that behave incorrectly because of this... > Ok this is clearly written in 3261, but you know how > developers do: "don't have time to read this long RFC, I will > just look at the example :p"... And the place you generally > look at is 3665, and then, it is broken (!)... > This is not "fixed" and I assume it will probably never be... > By the way this is a question: would it be interesting for > somebody to fix 3665 examples, and if yes, how to proceed? > + > Thomas > > > > > > -Jonathan R. > > > > Christer Holmberg wrote: > >> Hi, > >> > >> No matter how we'll fix this, I guess it would be a valid > essential > >> correction? > >> > >> Regards, > >> > >> Christer > >> > >> ________________________________ > >> > >> Lähettäjä: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] > >> Lähetetty: to 6.12.2007 21:09 > >> Vastaanottaja: [email protected] > >> Aihe: Re: [Sip] SIPit21: BNF future-proofing problem? > >> > >> > >> > >> From: Jonathan Rosenberg <[EMAIL PROTECTED]> > >> > >> This strikes me as a problem in theory and not in practice. > >> AFAICT, the > >> only characters you can't use are ascii 0x00-0x19 (0x20 > is space > >> and is > >> possible through the LWS construct as Christer pointed out). Is > >> there a > >> practical use case for any of these? They are: > >> > >> HT (horizontal tab) should be included, but that's also in > LWS. As > >> for the remaining characters, I don't think that anyone has ever > >> *intended* that they be used in SIP headers, and their > inclusion in > >> quoted-pair was a typographical error. > >> > >> Dale > >> > >> > >> _______________________________________________ > >> Sip mailing list https://www1.ietf.org/mailman/listinfo/sip > >> This list is for NEW development of the core SIP Protocol Use > >> [EMAIL PROTECTED] for questions on current sip Use > >> [EMAIL PROTECTED] for new developments on the application of sip > >> > >> > >> > >> > >> _______________________________________________ > >> Sip mailing list https://www1.ietf.org/mailman/listinfo/sip > >> This list is for NEW development of the core SIP Protocol Use > >> [EMAIL PROTECTED] for questions on current sip Use > >> [EMAIL PROTECTED] for new developments on the application of sip > >> > > > > _______________________________________________ Sip mailing list https://www1.ietf.org/mailman/listinfo/sip This list is for NEW development of the core SIP Protocol Use [EMAIL PROTECTED] for questions on current sip Use [EMAIL PROTECTED] for new developments on the application of sip
